Trata Verified
Fintech, Regulatory Technology (RegTech), Artificial Intelligence
Research Scientist Intern
San Francisco, California, United StatesOnsiteInternshipJunior / Entry-level$10,000–$17,000 /yrPosted 27 days agoHidden Gem · YC Startup
Role summary
Trata is seeking a Research Scientist Intern for the summer, with potential for contract work post-graduation. The intern will design datasets and evaluation rubrics to influence frontier model learning, with the opportunity to co-author a novel paper. Responsibilities include identifying model failure modes, building evaluation systems for RLHF/RLVR, and developing quantitative frameworks for dataset quality and impact. The role requires a deep curiosity about data's influence on model behavior, an ability to design experiments, and a bias for building. Prior experience in AI safety, benchmarking, or RL environments is a plus. Strong performers may be considered for full-time roles.
Trata is hiring a Research Scientist Intern for the summer. If post-grad, contract work is also an option. We are creating very unique benchmarks and evaluations. You will design datasets and evaluation rubrics that directly influence how frontier models learn. You will have the opportunity to co-write a paper that is largely the first of its kind. Your output will feed directly into model training runs at scale. This role requires facetime in NYC and SF (we pay for travel) but can include a remote portion for the right applicant.
What You'll Do
• Design data slices that expose meaningful model failure modes
• Build and refine evaluation rubrics and reward signals for RLHF and RLVR training pipelines
• Develop quantitative frameworks for measuring dataset quality, diversity, and downstream impact on alignment and capability
• Partner with lab research teams to translate training objectives into concrete data and evaluation specifications
What We’re Looking For
• Deep curiosity and motivation around how data structure, selection, and quality drive model behavior
• Ability to design lightweight experiments, move fast, and extract actionable insights from messy results
• A bias toward building over theorizing
• Prior experience at an AI safety org, benchmarking org, or RL environment company is a strong plus
Strong performers will be considered for a full-time offer.
To Apply
Send a brief note and your CV Links to evals you've built, experiments you've run, or writing on post-training are strongly encouraged.
About Trata:
Trata is backed by a VC with a higher unicorn rate than YC and Sequoia, Walter Kortschak (prev. CIO of SignalFire and Managing Partner at Summit Partners), Y Combinator, later-stage founders backed by some of the best VCs (Sequoia, Founders Fund, a16z, Accel, Bain Capital Ventures, YC), Pioneer Fund and 10+ hedge funds.
If interested, please follow on X: @trytrata
What You'll Do
• Design data slices that expose meaningful model failure modes
• Build and refine evaluation rubrics and reward signals for RLHF and RLVR training pipelines
• Develop quantitative frameworks for measuring dataset quality, diversity, and downstream impact on alignment and capability
• Partner with lab research teams to translate training objectives into concrete data and evaluation specifications
What We’re Looking For
• Deep curiosity and motivation around how data structure, selection, and quality drive model behavior
• Ability to design lightweight experiments, move fast, and extract actionable insights from messy results
• A bias toward building over theorizing
• Prior experience at an AI safety org, benchmarking org, or RL environment company is a strong plus
Strong performers will be considered for a full-time offer.
To Apply
Send a brief note and your CV Links to evals you've built, experiments you've run, or writing on post-training are strongly encouraged.
About Trata:
Trata is backed by a VC with a higher unicorn rate than YC and Sequoia, Walter Kortschak (prev. CIO of SignalFire and Managing Partner at Summit Partners), Y Combinator, later-stage founders backed by some of the best VCs (Sequoia, Founders Fund, a16z, Accel, Bain Capital Ventures, YC), Pioneer Fund and 10+ hedge funds.
If interested, please follow on X: @trytrata
Similar roles
Senior Research ScientistUniversity of Pittsburgh · Pittsburgh, Pennsylvania, United States · Onsite
Chief Research ScientistGreylock Partners · Redwood City, California, United States · Hybrid
Research ScientistCNA Corporation · Arrington, Virginia, United States · Onsite
Research ScientistVirginia Tech · Blacksburg, Virginia, United States · Onsite- Research Scientist Level 1Univeristy Of Colorado Boulder · Boulder, Colorado, United States · Onsite
